As a consequence of widespread screening mammography, approximately 20% of breast cancer diagnosed today is breast carcinoma in situ (BCIS). To reduce the chance of a subsequent breast cancer diagnosis after BCIS, many patients now receive treatment as aggressive as that recommended for invasive breast cancer. Several issues argue for critical examination of the impact of current BCIS detection and treatment practices. These include uncertain risk factors, substantial treatment morbidity, low numbers of BCIS cases that subsequently develop invasive breast cancer, negative impact on quality of life, and importantly, excellent relative survival regardless of therapy choice. A population predictive model, such as the one we have developed for breast cancer epidemiology, can accommodate the multi-dimensional nature of breast cancer. We propose a formal and quantitative evaluation of the uncertainties surrounding BCIS that incorporates multi- level risk factor and screening data into an epidemiologic study and an integrated simulation model assessing the population burden of BCIS. Specifically, we aim to comprehensively examine individual- and geographic- level socioeconomic and screening factors in relation to BCIS disease-free survival and treatment (among cases) and risk (comparing cases and controls). We further aim to use simulation modeling to evaluate the effect of BCIS on breast cancer incidence and outcomes. To accomplish these aims, we will continue to actively follow a large, population-based cohort of BCIS women, comprised of 2,352 cases, for disease-free survival. BCIS cases will be recontacted to ascertain any new breast cancer diagnoses and to evaluate health status and risk factor changes as measured by validated questionnaires. Available data for over 5,000 healthy population controls previously enrolled for our case-control studies will be included in case-control analyses jointly considering individual-level and geographic factors. Using epidemiologic data from the BCIS cases and controls as inputs, we will enhance our validated breast cancer discrete-event simulation model with individual- level risk factors, geographic measures of screening and treatment, and tumor histology. We will use the model to assess the contribution of BCIS diagnosis and treatment towards recent declines in breast cancer incidence and mortality rates since the 1990s, and to evaluate quality-adjusted life years resulting from screening and treatment regimens. Records from Wisconsin's statewide tumor registry, interview information, and publicly available data linked to the BCIS cases and controls through their geocoded residential addresses will contribute to both the epidemiologic cohort and case-control analyses and the simulation model. Using epidemiologic studies and computer modeling, this study will address BCIS uncertainties by characterizing the multitude of factors related to BCIS incidence, choice of treatment, quality of life, disease-free survival, and mortality. This study will efficiently and validly provide broad perspectives on BCIS, an increasingly important public health concern. PUBLIC HEALTH RELEVANCE: Women who are diagnosed with breast carcinoma in situ (BCIS) - 20% of all breast cancer diagnoses - have an excellent prognosis regardless of the type of therapy they receive. However, since women with BCIS are often treated with aggressive therapy and suffer similar side effects as women with more invasive cancer, the overall impact of BCIS detection and treatment on length and quality of life is not clear. The long-term goal of this project is to reduce the public health impact of BCIS by quantifying both contextual and personal risk factors for the development of BCIS and the recurrence of breast cancer after a BCIS diagnosis.